Quick Determination of Soil Quality Using Portable Spectroscopy and Efficient Multivariate Techniques

Rapid and onsite determination of the soil status quality parameters holds a brighter potential for improving food security, minimizing waste excessive application amendments hence reducing environmental pollution. In this study, pocket-sized shortwave NIR spectroscopy (740–1070 nm) multivariate statistics were used to classify from different land-use types simultaneously predict nitrogen (N), phosphorus (P), potassium (K), calcium (Ca2+), magnesium (Mg2+), pH in Ghana. Different Algorithms. Linear discriminant analysis (LDA), support vector machine (SVM), partial least squares algorithms (full-range square, FrPLS; interval squares, IPLS; synergy Si-PLS) attempted building suitable classification quantification model. The models assessed by rate, coefficient (Rp2), root mean square error prediction (RMSEP) set. A total 110 samples 0 15 cm, 30 45 cm layers collected field cropping systems. results obtained showed that SVM had 98.61% rate system. While Si-PLS was superior predicting N, P, K, Mg2+, Ca2+, pH. performance model 0.571, 0.779, 0.910, 0.778, 0.826, 0.904 Rp2 0.033%, 0.738 mg·kg−1, 0.117 cmol·kg−1, 0.654 3.0219 0.4760 unit RMSEP, respectively. revealed portable spectroscopic technique could be measure some parameters. However, further studies are needed proof its application. This lead yield saving cost fertilizer